> Date: Tue, 5 Mar 1996 15:41:14 +0000 (GMT)
> From: "S.J.Petrie" <email@example.com>
> One theory said that language
> was a 'spandrel'. A spandrel is the by product of another function; it
> wasn't deliberate.
In architecture, whence the analogy came, the circles on the Dome of
San Marco were deliberate (Michelangelo, the artist, deliberately drew
them); the spandrels were a byproduct. But in evolution, NOTHING is
deliberate: It is all just random genetic variation and selective retention
of the genes that breed success.
> An example of such a spandrel can be illustrated in
> the reason for female humans having two breasts. It is common for a
> woman to have one child but less frequent for her to have twins. That
> is why two breasts exist, ready for young to feed from. The presence
> of two breasts are there for the less predictable outcome ,even so.
> But what about if a female has triplets? One baby would have to wait
> until its sisters or brothers have finished. The other reason for
> having two breasts also comes under the law of balance and symmetry.
> One breast couldn't be physically possible; it would be too unstable
> for the woman to stand up straight.
That there are two breasts rather than one is a "spandrel" of
left-right symmetry of the vertebrate body; that there are not
six or eight, as with other mammals, is probably because primates
tend to have singleton births (because it takes a lot of effort to rear
one primate at a time, so there was selection AGAINST multiple births);
hence twinning and triplets, etc. are rare, and not "encouraged."
> It was thought that universal grammar was a spandrel. Except, what
> other use could it possibly be used for apart from communicating
> information in a way that everyone can understand?
Not clear what you mean here: Are you asking why all that detailed
structure was needed just for a mutual understanding protocol, to
make sure all communicators use the same conventions?
> The only theory of
> UG is that it is innate (inborn) which can't be explained as to why it
> happens just like the mystery of the big bang theory.
Not quite; the point is that if UG is not learnable and didn't evolve
gradually (how could it?) and doesn't make sense as a "spandrel" (of
what?) then the only candidate left is the Big Bang: that it was in the
nature of matter since the Big Bang that once you get neural tissue
gathered in a volume the size of a football, UG would be one of its
properties, just as the ring of 8 electrons is one of the properties of
the carbon atom...
> The poverty of the stimulus theory seems to explain UG better. This
> theory explains how children under four come out with completely
> acceptable phrases like 'he runned' as it follows the rules of regular
> verbs. However, the child can't have heard this from any source because
> adults just don't say it; it's not correct. There is simply not enough
> stimuli to learn incorrect structures. Even so, the child still
> produces new correct structures which haven't been taught. Somehow, the
> kid knows how to string together correct sentences without learning
> them through some kind of manual.
You need to read Pinker's other paper on language acquisition because
you have not yet understood this: That you do not see "he runned" but
"he ran" IS learnable, and IS learned. It is not a property of UG. UG is
much more abstract.
And "poverty of the stimulus" is not really a theory; it is more like a
datum (evidence, though it is disputed) plus an argument, to the effect
that the child never gets the negative examples he would need to learn
UG by age 4.
> Another matter of language is the arbitrariness of the sign, this being
> that something that stands for something else is known by everyone. Is
> that right?
No, arbitrary signs are signs that stand for things but do not resemble
them and are not causally connected to them. If the "name" of an apple
were a picture of an apple, that would not be arbitrary because it
resembled the apple. If the "name" were the taste of the apple, it would
not be arbitrary because it was causally connected to (biting into) the
apple. The sound "apple" (or the letters A P P L E), on the other hand,
are arbitrary names for apple, because they neither resemble nor are
causally connected to apples apart from the arbitrary convention English
speakers have agreed on, to call apples "apples").
> Something is nonarbitrary when an object is used to
> represent itself.
That's the most extreme way a sign (name) can be nonarbitrary, because
nothing resembles an object (or is more closely connected with it
causally) than itself. But pictures and other cause/effect connections
are also nonarbitrary.
> This object may have specific features that others
> wouldn't know, in this way it is nonarbitrary.
No, that has nothing to do with it...
> Language works with the
> issue of the arbitrariness of the sign. Words strung together are used
> to communicate a meaning.
Yes, but why would this not work with NONarbitrary signs: Pantomime,
for example. Would that be language? If not, why not?
> If we had no knowledge of what 'things'mean
> then language wouldn't be able to communicate them.
Kid-sib says that sentence did not say anything...
> The fact that
> everyone has a concept of what things mean allow language to strive by
> universal knowledge.
Same for this sentence. What do you mean? Yes, we all agree on what to
call things, otherwise we couldn't talk. But what other point are you
> Language is needed for successful communication but is grammar?
Some grammar, no doubt. But is UG needed, rather than something much
> Pantomime copes without stringing words together in no obvious rule; it
> relies on mimicking (copying others' actions). This is an iconic
> strategy, this then leads to symbolism when something represents
> (stands for) something else. Symbolism then leads to a functional
> advantage of acting out ,to saying words.
This sounds like it's going in the right direction, but kid-sib doesn't
get it: First things are "expressed" through pantomime and mimicry.
Fine. They become symbolic when the signs become arbitrary. Fine. But
what are the advantages of the latter, and the limitations of the
former. (Hint: Could we have had this conversation in pantomime? Why
> Mimicking can be passed on culturally rather than biologically.
Fine, but so can arbitrary symbols, so nothing depends on this...
> One generation teaches the next to act something out.
They also teach the next how to speak, and what to call what (with the
arbitrary, nonmicking vocabulary of a natural language)
> For example, rabbits
> learn to thump the grass when something threatening like a fox endangers
> theirs and others' lives. Other rabbits show what this means by
> previous experience.
You didn't quite get the point there: Rabbit thump, say, because they
are startled. But the thump startles their offspring too, sometimes, if
they thump hard enough: If this tended to save the offspring from
danger, and the tendency to do it varied (by chance) genetically, then
the next generation would thump harder...
This archive was generated by hypermail 2b30 : Tue Feb 13 2001 - 16:24:16 GMT